Pd2N2, a proteiform molecule: Matrix isolation spectroscopy and density functional theory calculations

The formation of Pd2N2 from the cocondensation of effusive beams of Pd and N-2 in neon and argon matrices is evidenced by absorptions in the range of 2200-1800 cm(-1). In argon, selective irradiation in the near-infrared and visible ranges leads to interconversions between three structures, distinguished by the stretching frequency of the diatomic N-2: Bridged T-shaped (nu(NN) at 1990 cm(-1)), side on (nu(NN) at 2178 cm(-1)) and parallel (nu(NN) at 1823 cm(-1)). For the first two structures, the nu(NN) mode is also accompanied by another signal below 500 cm(-1). An extra feature close to 490 cm(-1), not sensitive to irradiation at lambda(irr)>400 nm and also assignable to a molecule with the same Pd:N-2 stoichiometry (2:1), corresponds to a centrosymmetrical Pd-N-N-Pd structure with an inactive nu(NN) mode close to 2141 cm(-1), as deduced from the observation of a weak signal close to 2630 cm(-1) associated with this species and assignable to the combination nu(NN)+nu(PdN). All these experimental data and their structural implications are fully supported by theoretical calculations [density functional theory (DFT)]. On the basis of this comparative study, we have obtained a reliable theoretical description of the spectroscopic data using the metageneralized gradient approximation functional within the unrestricted DFT (UDFT) formalism for all spin multiplets. We have also searched a stable electronic solution for each multiplet (particularly for the singlet state), in order to account for the nondynamic correlations.